Natural Variation Identifies ICARUS1, a Universal Gene Required for Cell Proliferation and Growth at High Temperatures in Arabidopsis thaliana

Abstract

Plants are highly sensitive to environmental changes and even small variations in ambient temperature have severe consequences on their growth and development. Temperature affects multiple aspects of plant development, but the processes and mechanisms underlying thermo-sensitive growth responses are mostly unknown. Here we exploit natural variation in Arabidopsis thaliana to identify and characterize novel components and processes mediating thermo-sensitive growth responses in plants. Phenotypic screening of wild accessions identified several strains displaying pleiotropic growth defects, at cellular and organism levels, specifically at high ambient temperatures. Positional cloning and characterization of the underlying gene revealed that ICARUS1 (ICA1), which encodes a protein of the tRNAHis guanylyl transferase (Thg1) superfamily, is required for plant growth at high temperatures. Transcriptome and gene marker analyses together with DNA content measurements show that ICA1 loss-of-function results in down regulation of cell cycle associated genes at high temperatures, which is linked with a block in G2/M transition and endoreduplication. In addition, plants with mutations in ICA1 show enhanced sensitivity to DNA damage. Characterization of additional strains that carry lesions in ICA1, but display normal growth, shows that alternative splicing is likely to alleviate the deleterious effects of some natural mutations. Furthermore, analyses of worldwide and regional collections of natural accessions indicate that ICA1 loss-of-function has arisen several times independently, and that these occur at high frequency in some local populations. Overall our results suggest that ICA1-mediated-modulation of fundamental processes such as tRNAHis maturation, modify plant growth responses to temperature changes in a quantitative and reversible manner, in natural populations.

Fig 1. ICA1 growth phenotypes in natural accessions of Arabidopsis depend on temperature.

(A) 2 week-old Sij-4 and Don-0 plants grown at 23°C and 27°C. (B & C) Scanning electron microscopy of Sij-4 grown at 23°C and 27°C. Differences in cell morphology are visible upon high magnification (C) from day 10. Scale bar = 50μM. (D) Hypocotyl elongation at 23°C and 27°C under short days (SD) in Col-0, Sij-4 and F₁ plants derived from a cross between these two strains. 15–30 plants per genotype were analyzed for hypocotyl length measurement. *: p<0.0001; ns: not significant in Student t-tests comparing growth at the two temperatures. (E) Reversibility of the growth phenotype of Don-0 adult plants (4 week-old) in temperature shift experiments between 21 and 28°C. White arrows indicate new leaves developed after temperature shift.

Fig 2. Positional cloning of ICA1.

(A) Whole genome scanning of F₂ (Sij-4 x Col-0) along with marker positions and recombination frequencies (R.f.) with ICA1. Map positions are given in Mb. (B) Fine mapping of ICA1 in F₂ (Don-0 x Ler-0), F₂ (Sij-4 x Col-0) and F₂ (Sij-4 x Ler-0) segregating populations. Map positions and the number of recombinants are indicated above and below the line respectively. An overlapping interval of 5.9 kb contains a single gene. The single nucleotide deletion of Don-0 and the missense mutations between Col-0 and Sij-4 along with the corresponding amino acid changes are given below. (C) Phenotype of the Col-0 and ica1-2 (Col-0 background) grown at 27°C. (D) Phenotype of 35S::amiR-ICA1 in Col-0 background at 23°C and 27°C. (E) Phenotypic complementation of Don-0 with an ICA1-Ler transgene at 28°C (4 week-old plants). (F) Transgenic suppression of ICA1-Sij-4 phenotype at 27°C by 35S::ICA1gDNA-Col and lack of phenotypic suppression with the S81P mutation.

Fig 3. ICA1 affects cell cycle and endoreduplication.

(A) Expression pattern of Cyclin B1;1 in leaves of ICA1-Sij-4 and ICA1-Col plants segregating in an F₂ (Sij-4 x Col-0) population grown at 27°C and analyzed using a pCycB1;1::CycB1;1-GFP marker. (B) Shape of abaxial epidermal cells of the first leaf from Col-0 and Sij-4 plants grown at 27°C and visualized using the plasma membrane marker (pATML1::mCitrine-RCI2A). Magnification: left 10X, right 40X. (C) Epidermal cell nuclei of the first leaf of ICA1-Sij-4 and ICA1-Ler plants selected in an F₂ (Sij-4 x Ler) family grown at 27°C and visualized with the histone H2B marker (pATML1::H2B-mYFP). Magnification: left 10X, right 40X. (D) Quantification of nuclei sizes measured as mean (± standard deviation) nuclei area in ICA1-Sij-4 and ICA1-Ler-0 plants selected in an F₂ (Sij-4 x Ler-0) family grown at 27°C. *: p<0.0001 in Student t-test comparing nuclei sizes of both genotypic classes. (E) Flow cytometry analysis of Sij-4 plants compared with Col-0 and 35S::ICA1gDNA-Col in Sij-4 background at 27°C. Scale bars in A, B and C are 100μM.

Fig 4. Sij-4 plants are hypersensitive to DNA damage.

The sensitivity to DNA damage was assessed by the emergence of first leaves after Bleomycin treatment, which introduces double strand DNA breaks. (A) First leaf development in 10-day-old Col-0 seedlings grown at 23°C under long days. Examples for Sij-4 plants with (i) and without (ii) true first leaves are shown. (B) Percentage of plants developing true leaves when treated with Bleomycin (BLEO) or mock treated. Results from three independent experiments are shown for plants of Col-0, Sij-4, 35S::ICA1gDNA-Col in Sij-4 background and 35S::amiR-ICA1 in Col-0 background. Error bars indicate standard deviation from three biological replicates with 50 to 100 seedlings each. *: p<0.0001 in Student t-tests comparing pairs of genotypes.

Fig 5. ICA1 allelic variation and intragenic suppression by alternative splicing.

(A) Frc-0 displays the growth defect at 27°C and the phenotype of F₁ (Frc-0 x Don-0) plants demonstrates that ICA1-Frc-0 is another ICA1 loss-of-function allele. (B) Absence of the growth defects in Petro-1 and Dobra-1 accessions grown at 27°C. (C) Natural polymorphisms of major effect observed in ICA1 and their genomic positions. (D) Proportions of alternatively spliced transcripts of ICA1 at 23°C and 27°C in Col-0 and at 27°C in Dobra-1 and Petro-1. (E) Schematic representation of the three major splice forms (SF1, SF2 and SF3), along with the predicted stop codons (shown as red lines) in Col-0 and Dobra-1/Petro-1. The single nucleotide deletion of Dobra-1/Petro-1 is shown as a yellow line, whereas the region affected by alternative splicing is marked in orange color. The predicted protein lengths are indicated in the right side of panel. SF2 is due to an alternative splice acceptor site for I8 in E9 resulting in a shorter E9 exon; SF3 is due to an alternative splice acceptor site in the first intron resulting in a partial intron retention (IR).